The Role of Digital in the COVID-19 Social Assistance Response The Role of Digital in the COVID-19 Social Assistance Response | 1 © 2022 International Bank for Reconstruction and Development / The World Bank 1818 H Street NW Washington DC 20433 Telephone: 202-473-1000 Internet: www.worldbank.org This work is a product of the staff of The World Bank with external contributions. The findings, interpretations, and conclusions expressed in this work do not necessarily reflect the views of The World Bank, its Board of Executive Directors, or the governments they represent. The World Bank does not guarantee the accuracy of the data included in this work. The boundaries, colors, denominations, and other information shown on any map in this work do not imply any judgment on the part of The World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries. Rights and Permissions The material in this work is subject to copyright. Because the World Bank encourages dissemination of its knowledge, this work may be reproduced, in whole or in part, for noncommercial purposes as long as full attribution to this work is given. Any queries on rights and licenses, including subsidiary rights, should be addressed to World Bank Publications, The World Bank Group, 1818 H Street NW, Washington, DC 20433, USA; fax: 202-522-2625; e-mail: pubrights@worldbank.org. Cover photo: i_am_zews/Shutterstock.com 2 | The Role of Digital in the COVID-19 Social Assistance Response Contents Acknowledgements 4 Executive Summary 5 1. Context 7 Registration, Assessment and Eligibility Determination 2.  10 2.1. Registration and Assessment 11 2.2. Determining Eligibility 13 3. Payments 15 3.1. Delivering emergency response payments digitally 20 3.2. Building pathways to long-term development outcomes 23 4. Conclusion 27 5. Annex. Data and Methodology 29 Endnotes 31 Acknowledgements This paper was prepared by Georgina Marin (G2Px Program Officer) and Robert Palacios (Lead Social Protection Specialist) under the supervision of Vyjayanti Desai (G2Px Practice Manager) as part of the G2Px initiative, the World Bank Group’s cross-sectoral effort to support the digitization of government-to- person (G2P) payments in a way that accelerates financial inclusion, women’s economic empowerment and government-wide benefits. This paper benefitted greatly from the research assistance and inputs by Chrissy Martin Meier (G2Px consultant) and Ana Quiroz (G2Px consultant), and reviews by Harish Natarajan (Lead Financial Sector Specialist), Ines Rodriguez Caillava (Social Protection Specialist) and Nilima Chhabilal Ramteke (Senior Financial Sector Specialist). The paper builds on previous work from G2Px initiative working group and the broader World Bank Group, in particular those documented in “Social Protection and Jobs Responses to COVID-19: A Real-Time Review of Country Measures,” and data collected in partnership with the World Bank’s ASPIRE and JobsWatch teams. 4 | The Role of Digital in the COVID-19 Social Assistance Response Executive Summary The COVID-19 pandemic resulted in vast numbers of people in need of social assistance, many of whom were not previously covered by social safety nets. To meet this unprecedented level of need, governments quickly scaled social assistance reaching over 1.7 billion people in low- and middle- income countries. Scaling up social assistance presented two separate but related challenges: first, adapting targeting and registration to reach individuals not commonly included in social assistance databases, such as urban informal workers, and second, how to deliver government to person (G2P) payments safely and securely in the context of the pandemic. Countries that could leverage pre-pandemic investments in digital public infrastructure (DPI)— identification (ID), payments and trusted data sharing—were better able to implement COVID-response social assistance programs and reach more beneficiaries. This paper, analyzes the role of these DPIs, also called digital stack, in the social protection response to COVID by analyzing data on how COVID-response social assistance programs registered and made payments to beneficiaries across 178 programs across 85 countries. The analysis shows how these digital systems and infrastructure allowed for innovative targeting, registration, and payment approaches that covered a significant portion of the population. To tackle the first challenge, digital databases and digital ID systems played a critical role in the horizontal expansion of social assistance. Digitalized administrative databases and digital ID systems allowed for rapid cross-checking of databases to determine eligibility. Furthermore, in many countries, the use of the unique ID number and existing databases to determine eligibility made it possible for a ‘digital service window’ and mass registration. Overall, among the countries analyzed, those that were able to use digital databases and trusted data sharing to identify beneficiaries reached on average 51 percent of their population, while countries that had to rely on collecting new information from beneficiaries and couldn’t rely on existing databases to cross-reference either, reached on average only 16 percent of their population. Once beneficiaries were registered and their eligibility was verified, countries faced the second challenge: delivering funds, quickly and safely. 80 percent of countries included in this analysis had started to use digital payments for the delivery of at least one of their new or expanded social assistance programs as of May 2021. In total, an estimated 641.9 million recipients (218.3 million excluding China and India) received a digital payment –a huge step forward for financial inclusion. However, not all digital payments used for G2P delivery provided a gateway to deepen financial inclusion. Countries that invested in expanding financial inclusion pre-pandemic were better placed to leverage the more developed ecosystem to quickly and safely deliver payments. Being able to leverage existing accounts to deliver payments not only made the process easier for beneficiaries, it The Role of Digital in the COVID-19 Social Assistance Response | 5 also meant payments could reach them faster. Having vast access points for beneficiaries to cash-out or use their account were also key to a better recipient experience. However, many countries did not have this enabling ecosystem in place, and opted for one-time-passwords that did not provide the beneficiary with an account. Even among countries using accounts, some programs would only offer limited purpose accounts that didn’t allow beneficiaries to save, make purchases or transfer funds. Countries that shifted to digital payments during the pandemic, even if partially, now have the opportunity to leverage that investment to facilitate a long-term shift to modern social assistance payments. The digitization of COVID-response programs has led to an increase in account ownership. This provides a pathway to increasing financial inclusion only if this momentum is leveraged to develop and sustain the necessary enablers. At least 62 countries have leveraged account-based transfers for their COVID-response social assistance programs to some extent. Many of them, are using accounts as their social assistance payment method for the first time. Yet, in many cases, these account-based payments have been adopted only for temporary, COVID-response programs. Unless conscious efforts are made by governments to adopt these account-based payments across other social assistance programs and government payment streams, there is a risk of reversing the important strides made in terms of building the ecosystem needed to deliver digital payments. 6 | The Role of Digital in the COVID-19 Social Assistance Response 1. Context The COVID-19 pandemic resulted in vast numbers of people in need of social assistance, many of whom were not previously covered by social safety nets. When the health crisis came to bear and governments put restrictions in place to mitigate the spread of the disease, many workers lost access to income and many small business owners were forced to cease operations. Social assistance had to scale quickly: vertically, to provide additional support to vulnerable and rural communities already targeted by many social assistance programs; and horizontally, to reach the newly vulnerable, often informal workers in urban areas who did not qualify for benefits through social insurance programs or an employer. To meet this unprecedented level of need, governments quickly scaled social assistance. By May 2021, an estimated 1.7 billion people living in low- and middle-income countries had received COVID- related social assistance payments.1 In most regions, more than half of these recipients had never before received a social assistance payment. FIGURE 1. Percent of Population Covered by COVID-19 Social Assistance Programs by Region 63% 64% 44% 41% 33% 10% EAP ECA LAC MNA SAR AFR Note: Weighted average of population coverage in 80 low- and middle-income countries that scaled up social assistance programs in response to COVID-19, excluding India and China. Percent of population covered represents individuals living in households which had received a COVID-response social assistance payment as of May 2021. The Role of Digital in the COVID-19 Social Assistance Response | 7 Scaling up social assistance presented two separate but related challenges: first, adapting targeting and registration to reach individuals not commonly included in social assistance databases, such as urban informal workers (horizontal scaling), and second, how to deliver government to person (G2P) payments safely and securely in the context of the pandemic. Countries that could leverage pre-pandemic investments in digital systems—ID, payments and trusted data sharing—were better able to implement COVID-response social assistance programs and reach more beneficiaries.2 These systems are also sometimes called digital public infrastructure (DPI), in reference to the front- and back-end systems—provided by the government or in partnership with the private sector—that serve as ‘rails’ that underpin digital transactions and connections for people, businesses, and governments, including service delivery and operations across the public and private sectors, and in this case supporting social safety nets. This paper, analyzes the role of these systems in the social protection response to COVID by analyzing data on how COVID-response social assistance programs registered and made payments to beneficiaries across 178 programs across 85 countries.3 The analysis shows how these digital systems and infrastructure allowed for innovative targeting, registration, and payment approaches that covered a significant portion of the population. Countries that were able to use digital databases and trusted data sharing to identify beneficiaries reached on average 51 percent of their population, while countries that had to rely on collecting new information from beneficiaries and couldn’t rely on existing databases to cross-reference either, reached on average only 16 percent of their population. In particular, countries that leveraged digital systems both for the registration and payment process,4 reached the highest proportion of the population—an average of 55 percent (figure 2, upper right quadrant). All these countries also had high ID coverage, which facilitated matching and cross-referencing across existing datasets, as well as account opening wherever these were used. Yet, several countries that did not have the data to scale up coverage of cash transfer schemes quickly were still able to make digital payments especially by leveraging recent investments and growth in the mobile money market, especially in Africa (figure 2, bottom right-hand quadrant). This paper uses administrative data on G2P registration and payment methods combined with anecdotal evidence from country case studies to show how pre-pandemic investments in digital databases, digital ID, and digital payments impacted countries’ abilities to reach new beneficiaries and deliver payments safely in the context of the pandemic response. It further details workaround solutions implemented by countries without these assets and infrastructure in place, and how some countries were able to expand their digital infrastructure even amidst the urgency of the crisis response. The analysis concludes with suggestions as to the impact that the social assistance response to COVID-19 can have on the future of social protection payments, in terms of inspiring investments in building and strengthening G2P ecosystems globally. Insights from this data provide an exceptional opportunity to understand the role of digital investments in social protection. With millions of individuals now part of the financial and social protection systems for the first time, these insights support the opportunity to better design programs and products for vulnerable individuals and households—programs that can both meet their immediate needs and strengthen pathways to continued wellbeing and resilience. 8 | The Role of Digital in the COVID-19 Social Assistance Response FIGURE 2. Use of Digital Assets (Digitized Databases and ID Records) and Digital Payments in COVID-response Social Assistance Programs (SIZE: NUMBER OF INDIVIDUALS LIVING IN HHS COVERED BY THE PROGRAM) Fiji Pakistan Guatemala Peru Malaysia Indonesia Tunisia Jordan Cambodia Tonga Mauritania Brazil Thailand Mauritius Gambia Panama Dom. Argentina Namibia Rep. Ecuador Djibouti Turkey El Salvador Bhutan Yes Morocco Zimbabwe Registration, Assesment & Elibility Paraguay South Use of digital databases or ID records Lesotho Costa Rica Africa Togo Bolivia Egypt Jamaica Colombia India Mongolia Iran Chile Uruguay Mozambique CIV Vietnam Sierra Leone Tanzania Niger Guinea Zambia China Honduras Uganda Algeria Haiti Mali Kenya Lebanon Rwanda Sri Lanka Sao Tome Ghana No Benin Malawi PHL (SAP 1) Nigeria Myanmar Madagascar Burkina Faso Bangladesh Timor-Leste Ethiopia Manual (physical cash) Digital Payments ID coverage: Above 75% of adults Below 75% of adults Type of account: Mobile money Note: Use of digital databases is defined as countries that used existing digitized datasets to target and register beneficiaries. Use of digital payments includes countries that used traditional accounts, mobile money accounts or one-time-passwords. These accounts can be limited or fully functional accounts. Countries having more than one social assistance program in response to COVID are defined as using digital payment methods if at least one of the programs does so. ID coverage is defined as the percentage of adults with an ID (Global Findex 2017), complemented by ID population coverage from administrative datasets (ID4D) in cases where Findex data was not available. For Togo, given the relevance of voter IDs to their Novissi program rollout and that the ID coverage drastically increased after 2017, we are using a coverage of 97 percent. Similarly, for Malawi, given the coverage changes since 2017, the ID4D coverage is used instead of Global Findex. Source: G2Px-ASPIRE, ID4D, Global Findex. The Role of Digital in the COVID-19 Social Assistance Response | 9 2. Registration, Assessment and Eligibility Determination The COVID-19 pandemic resulted in vast numbers of people in need of social assistance, many of whom were not previously eligible. Many workers, particularly those in the informal sector, were unable to work for at least some portion of the pandemic due to restrictions put in place to mitigate the spread of the disease. Many of these informal workers neither qualify for benefits through an employer, nor have the typical profile of social assistance recipients, which, prior to the pandemic, were not poor but fell into poverty. In many countries, the scaling-up of social assistance focused on reaching these informal workers, as well as other segments of the population who were newly poor or vulnerable as a result of the pandemic. Scaling up social assistance vertically—providing additional benefits to those already benefitting from social assistance—was relatively simple. However, scaling up social assistance horizontally to reach those that were newly eligible proved much more difficult. It required registering individuals and households and determining eligibility. In many countries, digital databases and digital ID systems played a critical role in the horizontal expansion of social assistance.5 Countries with existing digital administrative databases and digital ID systems with high population coverage used these to rapidly scale up assistance and reach more beneficiaries (figure 3). These two digital assets were used to match records across datasets and identify potential beneficiaries on the one hand, and also to authenticate or verify their identity throughout registration. On-line application processes, for example, almost always relied on a unique identifier from the foundational ID system and in most cases, checked for eligibility criteria through other databases. While there were some exceptions where local governments played a major role such as the Philippines and Timor-Leste, countries without these assets found it more difficult to implement large horizontal expansions. Figure 3 shows that social assistance coverage was generally lower in countries that collected new data to target and register new applicants, relatively higher in countries that used existing databases and highest in countries that had both existing databases and high coverage ID. There are some exceptions where high coverage ID was available but not used (teal and yellow triangles). However, overall, the data shows a strong correlation between use of existing database and digitized ID records to high levels of social protection coverage. The following two sections will further detail how pre-pandemic investments in digital administrative databases and digital ID systems allowed for the rapid data sharing necessary for countries to accomplish two main objectives: registering additional individuals or households and determining eligibility. 10 | The Role of Digital in the COVID-19 Social Assistance Response FIGURE 3. Coverage of COVID-19 Relief % of population living in a household receiving a transfer by income and use of existing digital assets (digitized ID records and databases) Coverage: % Pop living in HH receiving a transfer 100 Timor-Leste Mongolia Bolivia Iran Malaysia Sri Lanka PHL (SAP 1) 80 Guatemala Thailand Brazil Morocco Indonesia Peru Argentina Tonga Panama South Africa Costa Rica 60 Jamaica Mauritius Myanmar Colombia Pakistan India Paraguay Chile 40 Turkey Fiji Uruguay Togo Gambia Mauritania Zambia Vietnam Ecuador 20 Cambodia Jordan Lesotho Bhutan Mozambique Haiti Lebanon Malawi Djibouti Rwanda Algeria China Zimbabwe Honduras Niger 0 7 8 9 10 11 Ln (GDP per capita) Collected new data Coverage of digitized ID records: Used existing databases Above 75% of adult population Used existing databases and digitized ID records Below 75% of adult population 2.1. REGISTRATION AND ASSESSMENT Countries that registered new applicants/target groups, either in new COVID-response programs or in horizontal expansion of existing programs, used essentially four different approaches: (1) using existing social registries; (2) using other administrative datasets; (3) implementing on-line applications; and (4) collecting information at the local level. Some countries combined more than one method to scale up their social assistance programs. Indonesia for example, expanded coverage to all households in their social registry and then added millions more using an on-line application process. Cambodia used its social registry to increase coverage from 1 to 16 percent of the population and then expanded registration during the crisis with commune officials collecting data. Figure 4 shows the number of countries that used one or more of the four methods of horizontal expansion, where data were available. The last bar shows that 75 percent of countries used digital infrastructure and databases—including social registries, online applications or other existing databases—to assess needs and conditions and determine eligibility. The Role of Digital in the COVID-19 Social Assistance Response | 11 FIGURE 4. Registration Method for Horizontal Expansions in 48 Countries (2021) 36 24 11 10 5 Local Social registry Application Other existing Any digital government database Note: Brazil and Indonesia are counted as multiple modalities. Categories are described in the following paragraphs. The easiest type of expansion, available to those countries with existing digital assets (digitized ID records and digital databases), was to rely on an existing social registry. This was the case in Cambodia and Indonesia as mentioned above as well as in Colombia and Pakistan. Brazil and Sri Lanka expanded to households on waiting lists before the pandemic. Some countries refined this approach and applied criteria such as checking that the person did not work in the formal sector. Argentina and Peru were among the countries able to check this by comparing social registry and social insurance databases. In Pakistan, where social insurance coverage is very low, the filters included having a public sector worker in the household, vehicle and property ownership, and the amount spent on telephone bills. This approach was most useful when the social registry covered a large part of the population and the information on households was up-to-date. Problems arose with some of the more outdated and static social registries leading to attempts to supplement or update the data. Ecuador supplemented its incomplete social registry with geographic targeting using census data combined with mobile phone usage data. In the case of the Philippines, while coverage was high, data had not been updated in more than four years and as a result could not be used for the cash transfer expansion. Federated social registries, where information is pulled from different databases that tend to be updated more frequently, as in the case of Egypt and Turkey, are more agile in this regard. 12 | The Role of Digital in the COVID-19 Social Assistance Response In lieu of a social registry, some countries were able to tap alternative administrative data. Togo used a recently generated voter database that covered 95 percent of adults and happened to include information on occupation which the government used for targeting informal sector workers. Guatemala relied on electricity consumption data. India used data on more than 200 million low- income women for whom bank accounts had been opened in a financial inclusion drive. Namibia relied on the civil registry and checked against income tax records. To register new applicants, many countries implemented an online application or self-registration process. In countries with digital administrative databases and a unique ID, online applications were quickly cross-checked against these databases. For example, in Turkey, applicants applied using an online application (with a call-in number offered as a back-up option) using their national ID. Once eligibility was confirmed, the applicant received their payment. During the emergency response, more than one million applications were processed per week. In at least 11 countries with insufficient digital assets (digitized ID records and digital databases) or administrative data, it was necessary to implement data collection by local government officials. Data poor countries with weak identification systems were forced to collect information. This was almost always done by local governments. Examples include Bangladesh, Madagascar, and Myanmar, countries where social registries did not exist and where administrative data are scarce.6 In Bangladesh, for example, tens of thousands of applications were rejected because enumerators had incorrectly entered ID numbers. The Philippines, a middle-income country unlike the other examples, was compelled to use local government officials to collect data from millions of households due to the lack of an up-to-date social registry and inability to establish an effective mechanism to allow potential beneficiaries to apply. Timor-Leste is a rare case where data collection efforts leveraged the electoral ID, allowing basic filtering of formal sector workers resulting in efficient and near- universal coverage. 2.2. DETERMINING ELIGIBILITY Digital administrative databases and digital ID systems allowed for rapid cross-checking of databases to determine eligibility. If an individual is already participating in multiple programs and appears in different government databases that might be useful for determining their eligibility, to the extent that this identifier links these different databases, the government is in a better position to filter for different criteria. One of the more common types of cross check was to confirm that the individual was not in the formal sector by looking for her identifier in the social insurance database. In many countries, the use of the unique ID number and existing databases to determine eligibility made it possible for a ‘digital service window’ and mass registration. This occurred in countries including Thailand, South Africa, Namibia, Togo, Peru, Fiji, Brazil, and Pakistan, among others. In Thailand, for example, the online application required only the national ID number. The system then checked twenty digital databases linked by the ID number to confirm eligibility. In the spring of 2020 alone, Thailand approved around 23 million applications from informal sector workers and farmers— more than half of the working age population. The Role of Digital in the COVID-19 Social Assistance Response | 13 TABLE 1. Examples of Databases Used for Determining Eligibility in a Horizontal Expansion Type of database Country examples Social registry Brazil, Colombia, Indonesia, Pakistan CRVS/ID Namibia, Pakistan, South Africa Income tax/payroll Malaysia, Namibia, South Africa, Pakistan, Brazil, Bangladesh Department of home affairs/immigration South Africa Health insurance for the poor Cambodia, Morocco Other social insurance data Brazil, Bangladesh, Thailand, Timor-Leste Education/ student registries Bolivia Social assistance beneficiary registries Brazil, Namibia, South Africa, Thailand, Bangladesh Electricity consumption El Salvador, Guatemala Financial inclusion program data India Electoral registry (voter ID) Togo Mobile phone usage/billing data Togo, Pakistan, Bangladesh Farmer/ land registry data Thailand Source: Johnson and Palacios 2022. Data sharing arrangements supported both assessment of needs and conditions and eligibility determination, yet data privacy and protection were not always in place. On the one hand, countries such as Turkey, where the data sharing arrangements were governed by a data privacy law, made efforts to ensure consent from applicants. When an individual applied for social assistance, they received a list with each institution that would access their information, and every family member had to sign the form in order to provide consent. On the other hand, there were countries where data was not managed and shared in line with good practices for data protection and information security such as in the Philippines, where the lack of government data sharing infrastructure and low capacity at local government levels led to personal data being shared by email and cloud spreadsheets.7 Similarly, in Paraguay, Google drive spreadsheets were posted in the government website for everyone to see the list of eligible applicants including ID numbers and municipality. 14 | The Role of Digital in the COVID-19 Social Assistance Response 3. Payments Once beneficiaries were registered and their eligibility was verified, countries faced the second challenge: delivering funds, quickly and safely. Those countries with pre-pandemic investments in digital infrastructure turned to digital payments, as opposed to manual, over-the-counter cash payments, largely motivated by the desire to ensure the efficient and socially-distanced delivery of funds. Digital payment methods included account-based payments, where beneficiaries received the transfers directly into their traditional bank or mobile money accounts, and non-account digital methods such as one-time passwords (OTPs).8 80 percent of countries included in this analysis had started to use digital payments for the delivery of at least one of their new or expanded social assistance programs as of May 2021 (figure 5.1).9 In total, an estimated 641.9 million recipients (218.3 million excluding China and India) received a digital payment—a huge step forward for financial inclusion.10 If we focus only on countries that used account- FIGURE 5.1. Number of Countries by Payment Methods Used During the COVID-19 Response 44 1 Traditional + Mobile 19 money account Mobile 18 money 12 account 3 5 Traditional 17 bank 12 10 account 6 Account only Multiple payment Electronic Manual only methods, including non-account accounts only The Role of Digital in the COVID-19 Social Assistance Response | 15 FIGURE 5.2. Number of Recipients by Payment Methods Used During the COVID-19 Response 178.7m Traditional + Mobile money account Mobile money 59.3m account 44.8m Traditional bank 10.8m account Account only Multiple payment Electronic Manual only methods, including non-account accounts only Note: Data excludes China and India. based payments, we can estimate that at least 207.5 million direct recipients (excluding China and India) were beneficiaries of a program that used accounts as one of their payment delivery methods (figure 5.2).11 This means that at least 70 million recipients of COVID-response social assistance payments (in addition to many other regular programs that were not modified during COVID-19 and thus are not part of this analysis) are yet to receive payments through an account—a huge opportunity to expand financial inclusion and pursue efficiency gains. However, not all digital payments used for G2P delivery provided a gateway to deepen financial inclusion. Electronic non-account payments such as one-time passwords and single use codes were leveraged by some countries that were lacking essential enablers, such as the ability to conduct remote account onboarding, but still wanted to gain some of the efficiencies of delivering payments digitally. This payment method allows for simplified delivery with no onboarding needed, as an account is not created. Instead, a one-time password or a single-use code is either delivered or requested by the beneficiary and can be used to carry out purchases or withdraw cash at merchants or ATMs. As these payments are not linked to an account, this method does not provide a gateway to other financial services such as savings, transfers, and digital payments, which can deepen financial inclusion and improve beneficiaries’welfare. Among the 85 countries analyzed, at least six countries reported leveraging these electronic non-account-based methods, including Sierra Leone and Guatemala. While these methods don’t currently expand financial inclusion, if the program incorporates strategies to facilitate opening and linking of accounts, it is possible to leverage the unique-code based payments 16 | The Role of Digital in the COVID-19 Social Assistance Response as a pathway to receiving transfers directly into an account. For example, a recipient can receive a unique-code based payment in her phone and choose to either cash it out or upload it to her wallet or bank account. Even among programs using accounts, not all were equally conducive to financial inclusion. Among countries leveraging digital payments, at least 103 programs across 55 countries (58% of programs) used fully functional accounts either as the sole payment method or in conjunction with others12. In addition to these, 34 programs across 13 countries also leveraged accounts but limited the way in which they could be used.13 For example, in Paraguay, beneficiaries received payments via newly created mobile money accounts, yet the limited functionality of these accounts did not allow recipients to use the wallet to save or make transfers. Beneficiaries could only use unique-code based payments to cash out and make purchases. Countries leveraged mobile money accounts to deliver G2P payments more than ever, as they provided an option to reach beneficiaries even in ecosystems with lower traditional financial sector penetration. Among countries that used account-based payments, there were two types of accounts used: traditional bank accounts and mobile money accounts. For each country, the selection of account type was influenced by several factors, including the relative degree of each type of account penetration pre-pandemic and the available financial access points (figure 7.1 and 7.2). Countries that did not have high penetration of the traditional banking sector, but had invested pre-pandemic in the growth of mobile money, were able to use these investments to deliver account-based payments during the COVID-19 response (figure 6, pie charts, teal.) Many of these countries are in sub-Saharan Africa, where mobile money uptake is higher than that of traditional banks. Yet, several countries in Latin America and the Caribbean, including Paraguay, Peru, and Colombia, also used mobile money accounts in addition to traditional accounts to expand the reach of digital payments to new segments of the population (figure 6, pie charts, dark blue.) The Role of Digital in the COVID-19 Social Assistance Response | 17 FIGURE 6. Selection of Account Type for Payment Delivery Varied by Region East Asia and Middle East and Pacific (n=12) North Africa (n=10) Account only 50% Account only 20 20 Multiple payment methods, 50 % 50 % 40% 17% including accounts Multiple payment methods, including accounts 10% Electronic non-account only 0% 60 Electronic non-account only 10% Manual only 33% Latin America & Manual only 40% the Caribbean (n=22) 12.5 South Asia Account only 36% 37.5 % 50 Sub-Saharan Africa (n=6) Multiple payment methods, including accounts 36% (n=33) Account only 67% 20 Electronic non-account only 14% Account only 61% 19 Multiple payment methods, including accounts 17% 20 % 60 Manual only 14% Multiple payment methods, including accounts 18% 69 % 12 Electronic non-account only 0% Manual only 17% Electronic non-account only 6% Manual only 15% Traditional/Bank account Traditional & Mobile account Mobile accounts 18 | The Role of Digital in the COVID-19 Social Assistance Response FIGURE 7.1. Percent of Account Owners with a mobile Money Account in Countries Using Accounts to Deliver G2P by type of account used during COVID response (% of account owners) 55 15 19 Traditional Both types of Mobile money accounts accounts accounts Source: 2018 Global Findex and 2021 G2Px/ASPIRE administrative dataset. Note: Considers payment methods used in each country for COVID-response programs where payment information is available. FIGURE 7.2. Mobile Infrastructure and Brickand Mortar Financial Access Points by country payment method 2000 1500 Mobile money agents 1000 500 0 0 50 100 150 Bank acess points – branches, ATM and agents Traditional accounts Mobile accounts Manual Both types of account Electronic non-account Source: 2019 GSMA Mobile Connectivity Index, IMF Financial Access Survey using most recent data for each country, and 2021 G2Px/ASPIRE administrative dataset. Note: Bolivia and Mongolia are not included for presentation purposes given their high branch penetration. 19 | The Role of Digital in the COVID-19 Social Assistance Response 3.1. DELIVERING EMERGENCY RESPONSE PAYMENTS DIGITALLY Countries that were already using digital payment methods to deliver social assistance payments prior to the pandemic were in a much better position to leverage these for the COVID-19 response. From a sample of 50 developing countries for which we know at least one program prior to COVID-19 was using digital payments only a handful did not continue using digital payments for their new or expanded programs. In countries like Ecuador, for example, prior to COVID, beneficiaries were given the option to receive payment into their own account (although only about 15 percent of recipients were using this option). However, in the COVID-19 response program all payments were done manually (in cash over the counter) albeit using financial access points for delivery. In other countries, like Indonesia, where digitization was already adopted by most social assistance programs prior to the pandemic with most social assistance beneficiaries receiving payments through a state-owned bank account, the BLT Dana-Desa program targeting village residents not included in the social registry decided to deliver payments manually given the difficulties of opening accounts for beneficiaries and enabling cash-out in remote locations with low connectivity and low financial access points. What were some of the things that enabled countries to use digital payments to deliver COVID- response programs? First, beneficiaries had to be onboarded to an account or the information of their existing accounts needed to be registered by the program. Having a good ID system in place, having remote and simplified KYC regulations, and already high account penetration were useful in achieving this. Second, beneficiaries had to be able to cash out or use their benefit digitally. To achieve this, a sufficient financial access point network is needed, as is, ideally, widespread electronic payment acceptance among merchants in their ecosystem. Finally, underlying both of these critical steps of the beneficiary journey is the need to have payments system infrastructure and arrangements with payment service providers in place. Digital ID systems supported countries’ efforts to leverage digital payments. Having a digital ID not only facilitated eligibility determination in several countries by leveraging various datasets or online registration, they also enabled countries to quickly onboard beneficiaries into accounts. Accounts, whether traditional or mobile money accounts, necessitate verification of identity, as per customer due diligence (CDD) account opening requirements.14 The higher the ID coverage in the country, the more feasible it was to quickly on-board people into a new account amid the pandemic. Thus, pre- pandemic investments in digital ID systems (represented by higher coverage of digitized ID systems) were positively correlated with the use of digital payment methods (figure 8). In addition to high ID coverage, there were two policies that supported account opening during the pandemic: simplified CDD and remote (also called non-face-to-face) CDD. Alongside relatively high ID coverage, countries such as Peru and Colombia were able to leverage the simplified and remote know-your-customer (KYC) policies already in place prior to the pandemic, enabling beneficiaries to open accounts easily and remotely during the pandemic. Many of those without these pre-existing policies—at least 12 countries including Cote d’Ivoire, Mali, Togo, Thailand, and Jordan—made regulatory changes during the pandemic to ensure recipients could open accounts with fewer requirements and/ or remotely during the pandemic. In Jordan, for example, the Central Bank allowed all payment service providers (PSPs) to facilitate opening wallets online and with simplified KYC. In Cote d’Ivoire, Mali, and Togo (alongside all other members of the Economic and Monetary Community of West Africa), the requirements for opening electronic money accounts were relaxed, allowing individuals to do so with only their mobile number. This allowed countries to reach beneficiaries with mobile money accounts, even with lower ID coverage. The Role of Digital in the COVID-19 Social Assistance Response | 20 FIGURE 8.1. Higher Average Coverage of Digitized ID Systems Correlates with Digital Payments (of adults with a digitized ID) 77 66 Manual Digital N=85; Source: ID4D, Global Findex. For more information on digitized ID coverage variable construction, see note for Figure 2. FIGURE 8.2. Country has Simplified Customer due Diligence (CDD) Requirements, by G2P Payment Method Used (% of countries with or without policy) 59% 22% 13% 6% Yes 44% 15% 41% No Account only Multiple payment methods Electronic non-account Manual only N=59; Source: 2017 Global Financial Inclusion and Consumer Protection Survey and 2021 G2Px/ASPIRE administrative dataset. Note: Considers payment methods used in each country for COVID-response programs where payment information is available. 21 | The Role of Digital in the COVID-19 Social Assistance Response FIGURE 9. Average Account Ownership by Countries’ Payment Method (% of adult population, N=71) 60 54.4 45.4 41.0 40 36.9 20 0 Account only Multiple payment Electronic Manual only methods non-account Source: 2018 Global Findex and 2021 G2Px/ASPIRE administrative dataset. Note: Considers payment methods used in each country for COVID-response programs where payment information is available. Countries that invested in expanding financial inclusion pre-pandemic were able to use existing account penetration for the quick delivery of payments. Countries that used account-based payment methods to deliver COVID-response social assistance payments had, on average, higher account ownership penetration among their adult population than countries that used manual or electronic non-account-based methods (figure 9). This correlation could be explained by two factors: first, countries could use these existing accounts to deliver payments as was the case in Colombia, Thailand, and Peru; second, higher levels of account penetration also indicate that a country already had the legal and regulatory framework and market conditions in place to support digital payments, and therefore were also better able to quickly open new accounts for beneficiaries where needed. Being able to leverage existing accounts to deliver payments not only made the process easier for beneficiaries, it also meant payments could reach them faster. Beazley et al (2021) find that programs utilizing digital payment methods were implemented more quickly, on average, than other programs. Even within programs using digital payment methods, implementation was faster for those that were able to rely on existing accounts. For example, in Colombia, over 79 percent of adult Colombians had an account where they could receive deposits as of 2019.15 To scale up social assistance during the initial months of COVID-19, the state-owned bank, Banco de las Oportunidades, leveraged a data-sharing agreement with a credit information registry (to which all banks report) to identify if targeted households had at least one member with an account. From this exercise and following the confirmation of these accounts with all respective financial institutions, the Colombian government was able to identify accounts for over a third of the targeted households to which payments could be made as early as April 8, 2020.16 For the remaining households, a strategy was deployed in partnership with financial institutions to open mobile wallets through remote onboarding. Payments started The Role of Digital in the COVID-19 Social Assistance Response | 22 almost three weeks later, on April 27. Furthermore, leveraging existing accounts also translated into cost savings for the government, as the fees paid to PSPs for banked beneficiaries were almost half of those paid to onboard and pay new beneficiaries. Yet, sending the payment was just half the journey: beneficiaries still had to be able to use their benefits. In countries that used digital payment methods, the availability of cash-out points was a critical variable to support implementation. In these countries, the digital payment ecosystem, in general, does not yet support entirely transacting digitally due to relatively low penetration and acceptance of electronic payments, making cashing out a necessity for beneficiaries. On average, countries using digital payment methods had higher penetration of ATMs, commercial bank branches and agents per capita.17 Where bank access points penetration (branches, ATMs, and bank agents) was low, many countries were still able to leverage digital payments for delivery by leveraging their mobile money agent networks (see figure 7.2). In other cases, ad hoc solutions were crafted to reach remote beneficiaries with digital payments. In Angola and Liberia, where mobile money and traditional accounts were used to deliver payments,18 agents were deployed from urban centers to more remote locations on specific days because there is no agent network in several areas where the cash transfer program makes payments. In Liberia, 62 % of surveyed beneficiaries said it would be difficult to get their payments without these “cash out events.” In Brazil, boat agents were deployed monthly to 31 remote municipalities to reach over two thousand, mostly indigenous, beneficiaries that would otherwise need to travel long distances to cash out.19 Financial access points also supported non-account based digital payments. Countries like Guatemala, Honduras, and Morocco that used one time passwords (OTP) sent to mobile phones also leveraged financial access points for beneficiaries to cash out. In particular, beneficiaries could use the OTPs to cash out at an ATM,20 bank branch, or agent. This method helped to achieve the quick scale up of social assistance in these countries. However, this method did not support pathways to financial inclusion as no accounts were linked to these OTPs, thus not allowing for savings, sending of remittances to friends and family, or the building of a transaction history necessary to access other financial services including loans and insurance. Even in countries where payments were provided manually, such as Bolivia and Cambodia, financial access points were also leveraged as cash out points where beneficiaries could identify themselves and receive their payments without an OTP or an account. 3.2. BUILDING PATHWAYS TO LONG-TERM DEVELOPMENT OUTCOMES Most countries that used digital payments to deliver COVID-response payments had already implemented these to some extent prior to the pandemic; however, several countries used them for the first time, facilitating a long-term shift to modern social assistance payments. For example, Liberia and Cameroon leveraged mobile money accounts for the delivery of social assistance payments for the first time. In Angola, digital payments were also used for the first time, with some beneficiaries receiving their transfers into a mobile money account in their name, although others received their payment via an  ATM card  linked to a sub-account under the program  that could be used at any ATM  terminal  or Point of Sale (POS) terminal, but was not a fully functional account. Similarly, in Mozambique, payments were delivered using a mix of mobile money accounts and limited prepaid cards that could be used to cash out and make purchases, but did not allow deposits other than 23 | The Role of Digital in the COVID-19 Social Assistance Response from the program. Some other countries, such as Guatemala, Honduras, Malawi, Morocco, and Sierra Leone, adopted digital payments based on digital vouchers or one-time-passwords. While these had no linked account, beneficiaries could benefit from more widely available cash-out points than when receiving over-the-counter payments in government offices or at cash-out events. For example, in Guatemala beneficiaries could cash out from any ATM and a broad network of banks and agents. Countries that opened accounts for beneficiaries in order to make payments created new pathways to financial inclusion. While higher account ownership can be an enabler for G2P payment digitization it is also one of the desired outcomes of digitizing G2P payments. During COVID-19 many countries used digital payments for COVID-response programs—not only for its efficiency and convenience, but also to advance financial inclusion. In Jordan, the number of active e-wallets doubled in four months, reaching 1 million users.21 In Paraguay more than 1.5 million wallets were created to receive social protection payments.22 In Brazil, close to 70 million beneficiaries received aid through the Auxílio Emergencial program, for which the government set up a digital savings account with digital access channels to use the funds. It is estimated that 40% of beneficiaries did not have an account before the pandemic.23 In Colombia, almost all 3 million beneficiaries of Ingreso Solidario received their payments through an account, with over 1.3 million new mobile accounts opened among unbanked beneficiaries for this purpose.24 In Argentina, it is estimated that the recent opening of more than 1 million universal free accounts was related to the obligation of having a bank account to receive aid from the Ingreso Familiar de Emergencia program.25 In the Philippines, 11 million beneficiary households (out of 14 million) received the second disbursement of the Social Amelioration Program in a bank account or mobile wallet, including existing CCT beneficiaries receiving disbursement in their usual account and some who provided their own account information through an online platform. However, most had accounts or mobile wallets opened on their behalf by the government through a partnership with six financial service providers.26 Furthermore, without the ability verify account information in cases where this was provided, there were data quality issues, which led to delayed payments Not only did account ownership increase among beneficiaries, but also account usage. In Colombia, for example, of the 1.7 million beneficiaries from Ingreso Solidario that received their payment through a mobile money account, 22 percent also made deposits or received other payments into their accounts and 23 percent transferred money to peers.27 In Brazil, from the total funds transferred by the program Auxilio Emergencial to mobile money accounts, 75 percent was used digitally and only 25 percent was cashed out.28 There were explicit efforts from the Auxilio Emergencial program to increase the use of funds digitally, including not allowing beneficiaries to cash out immediately after the payment: instead there was a period of between 10 to 53 days in which funds could only be used to make transfers or digital payments.29 In some countries where account-based payments were already used, choice was introduced for the first time or considerably scaled up. Offering a choice of PSP to beneficiaries can potentially increase convenience for recipients, create efficiencies for government, and support the development of a more inclusive financial services market.30 PSP choice can be especially important in low interoperability contexts, where without choice the beneficiaries are bound to a PSP which might not have appropriate access channels or cash out networks near to their location. In Tunisia, for example, new beneficiaries could choose their type of account (postal, mobile money, or traditional bank account) as well as their provider, for the first time during COVID-19. In contrast, Jordan had already been investing in the modernization of G2P payments prior to COVID-19. During the pandemic, plans to digitize social The Role of Digital in the COVID-19 Social Assistance Response | 24 assistance payments were scaled up and beneficiaries received a mobile-based choice of PSP. Yet most programs continued to offer no choice, even when multiple PSPs were procured for the delivery. For example, Colombia assigned beneficiaries to a PSP using a pre-defined set; the program did not allow beneficiaries to choose among the providers. In the Philippines, the second round of the Social Amelioration Program that delivered the payments digitally leveraged various PSPs including banks and e-money issuers, but these were assigned geographically based on their access point networks with beneficiaries having no choice.31 While significant progress has been made in digitizing payments during the COVID-19 crisis, challenges to reaching long-term development goals such as financial inclusion, women’s economic empowerment, and government-wide efficiencies have surfaced: • Low beneficiary awareness and financial literacy became a barrier to full financial inclusion. For example, in the Philippines only 31% of surveyed beneficiaries who received digital payments under the SAP 2 COVID-response program could correctly recall which FSP disbursed their allowance, and only 16% reported that they have an account for their SAP allowance. Among those who reported having an account only 60% knew it had other uses.32 In Liberia, 81% of beneficiaries reported having never checked their mobile money balance, among which 96% did not even know how to check it. • In many countries, account design and policy rules limited the useability of the account and therefore the path to full financial inclusion. For example, in Paraguay, beneficiaries could not make deposits or receive transfers—other than the program’s—into their mobile money accounts. Similarly, in the Philippines, accounts opened for SAP 2 were limited-purpose accounts. These can be upgraded to fully functional accounts if beneficiaries complete a full CDD with the PSP in the future. However, the onus of initiating this process lies with the beneficiary. Account design is not the only barrier to full use: program policies that erode trust and incentivize beneficiaries to fully cash out their benefit such as clawback clauses are also a barrier. In Indonesia, for example, in programs that were using accounts the benefits are clawed back after a month if not fully withdrawn, reducing trust and usage. Similarly, in Brazil, the Auxilio Emergencial program clawed back the resources if these were not fully spent within 120 days.33 • Underdeveloped digital financial services (DFS) ecosystems, where electronic payments are not commonly accepted and peers do not use digital transfers, limited the use cases for beneficiaries, curtailing the potential benefits of digitization. For example, in Angola, Central African Republic, and Liberia beneficiaries have been documented to cash out fully and immediately because digital payment are rarely accepted in rural areas. Furthermore, in the case of Angola and Liberia, cash-out events are necessary in many regions as there are no financial access points, transforming the account into little more than an identity verification instrument. • While several countries directed payments to women, others missed the opportunity to build pathways to women’s economic empowerment, and some programs even excluded women due to their design. Many COVID-19 response programs directed digital payments to women (figure 10). Women, already at a disadvantage due to systemic gaps in access to digital, financial, and other services, were disproportionately impacted by the negative economic impacts of COVID-19. Globally, women took on more of the childcare burden imposed by school closures and were more likely to fall ill. Evidence shows that directing payments to a woman, into an 25 | The Role of Digital in the COVID-19 Social Assistance Response account in her own name, can empower her in the face of such obstacles.34 In India, for example, one COVID-response social assistance programs exclusively targeted women: transfers were made directly to over 206 million women in their existing PMJDY accounts.35 While the program faced challenges, including lack of awareness about the payment and PMJDY not having complete coverage across all lower income women, the program was responsive to the gender specific challenges women face by targeting them specifically.36 Other countries didn’t direct payments only to women but did implement policies to prioritize them. For example, in Brazil and Togo, the transfer amount was larger for female recipients. In Peru, the eldest woman in the household was prioritized as the recipient. However, as the program adopted digital payment methods, household members with existing bank accounts were prioritized over women who did not have an account. Similarly, in Cameroon, payments were initially directed to women in the household, but lack of IDs and mobile phones led to replacement of designated recipient. These experiences provide a cautionary tale for considering gender differences in access to IDs, bank accounts, and mobile phones when implementing digitization strategies. Finally, many countries defaulted to the head of household as recipient given social norms (Mali) and the difficulty to switch to a different household recipient, especially given the lack of female household member information (Jordan, Tunisia37). FIGURE 10. Examples of COVID-Response Programs Directing Payments to Women (% of adults with a digitized ID) Account-based Mixed 100% 100% Nutrition Sensitive Direct Support (NSDS) 100% 100% 86% Maternal and Child Cash Transfer Program Social Safety nets Project (SSNP) 80% 81% 71% 47% 65% Bono Familiar Universal 60% 58% 56% Social Cash Transfer (STC) 54% Emergency Cash Transfer Post-Emergency Social Assistance Program 40% Druk Gyalpo’s Bono Familia Relief Kidu Solidarité Fonds de Kwenda PMJDY Novissi PKH 4Ps India Indonesia Rwanda Philippines Angola Liberia Cameroon Togo Sierra Guatemala Côtes Myanmar Peru Mozambique Bhutan Leone d’Ivoire Digital non-account based Source: Greta Bull et al. 2021. “Empowering Women: Send Social Assistance Payments Directly to Her.” World Bank Blogs. March 5 2021. Note: For more information on the estimations and exceptions, please see original figure note here. The Role of Digital in the COVID-19 Social Assistance Response | 26 4. Conclusion Countries that made investments in digital public infrastructure (DPI)—digital ID, data sharing, and digital payment systems—pre-pandemic were able to leverage this to expand coverage at a higher scale and make payments more efficiently and safely through digital methods. This resulted in countries being able to enroll new beneficiaries with minimum to no in-person contact, beneficiaries having more convenient access to benefits, and in many cases the deepening of financial inclusion. Their experiences in doing so demonstrate the value of investments in these enablers in terms of setting the basis for responsive and resilient social protection, as well as for more efficient and convenient G2P payments more generally. At the same time, this analysis found that while all of these investments played an important role in facilitating the response, no single factor can be attributed to enabling quicker payments or higher coverage. In addition to the DPI, there were also policy changes that made the adoption of digital financial payments during the pandemic possible. These included simplified customer due diligence regulations that allowed beneficiaries to open accounts with fewer requirements as well as for remote account opening in some cases; the creation of basic or no-frills accounts to ensure a minimum set of services was available to recipients at low or no cost; and, the increase or removal of transaction limits as well as fee waivers or reductions to incentivize digital transactions. Thus, this analysis supports the need for a holistic and comprehensive approach to investing in the infrastructure, policies, and processes necessary for effective, efficient, recipient-centric and crisis-responsive G2P architecture.38 Countries that shifted to digital payments during the pandemic, even if partially, now have the opportunity to leverage that investment to facilitate a long-term shift to modern social assistance payments. The digitization of COVID-response programs has led to an increase in account ownership. This provides a pathway to increasing financial inclusion only if this momentum is leveraged to develop and sustain the necessary enablers. At least 62 countries have leveraged account-based transfers for their COVID-response social assistance programs to some extent. Many of them, are using accounts as their social assistance payment method for the first time. Yet, in many cases, these account-based payments have been adopted only for temporary, COVID-response programs. Unless conscious efforts are made by governments to adopt these account-based payments across other social assistance programs and government payment streams, there is a risk of reversing the important strides made in terms of building the ecosystem needed to deliver digital payments. G2P digitization required strong coordination among relevant stakeholders, and these efforts must be sustained to scale up digital G2P payments, as well as DPI more broadly. Coordination across government agencies (e.g. to leverage datasets across agencies to support eligibility assessment) and between the private and public sector (e.g. to identify account holders or to disburse the payments at low or no cost) was essential for the deployment of digital G2P payments. Digital G2P payments require a cross-sectoral, whole-of-government approach. It is important that this interministerial and public-private collaboration is maintained and that the G2P architecture that was created to deliver emergency payments is fully adopted by social protection agencies. This will require building the 27 | The Role of Digital in the COVID-19 Social Assistance Response necessary institutional capacity to support the ongoing digitization of G2P payments. Furthermore, it is critical that the collaboration across government and payment service providers is based on a sustainable business model that contributes to the further development of the digital financial ecosystem and accelerates financial inclusion. The opportunity to scale up the digitization of G2P payment streams beyond social assistance payments requires that G2P digitization is approached holistically. Implementing a modern G2P architecture39 that is based on shared infrastructure and that leverages the key identification, national payments, public financial management and social protection systems in the country can bring great efficiencies and benefits to both recipients and government. A modern G2P architecture, enabled by inclusive and trusted DPI, provides the shared rails for different G2P payment streams (across programs and agencies) to be delivered digitally, from end-to-end. This would not only increase government-wide fiscal savings but would also create an important volume of payments that can contribute to the development of the payments ecosystem in terms of financial access points and adequate financial products and services. This can ultimately contribute to increasing beneficiaries’ convenience, inclusion, and empowerment. The Role of Digital in the COVID-19 Social Assistance Response | 28 5. Annex. Data and Methodology This analysis draws from a new, cross-country dataset of COVID-19 cash transfer programs that provide comparable, country-level estimates of the share of individuals living in a household that received a COVID-19 cash transfer, estimations of the coverage of digital ID systems, and the payment methods used. To build this dataset, multiple databases and quantitative sources were leveraged to allow for the most thorough assessment to date on the digital aspects of the COVID-19 emergency response. Sources include: • The ASPIRE Administrative Data Platform, which, in collaboration with G2Px and the World Bank’s Global COVID-19 social protection database, collected registration and payment method information for 196 programs across 94 high income and developing countries; • The Social Protection Responses to COVID-19 in the Global South database compiled by the International Policy Center for Inclusive Growth (IPC); • Country specific reports from publicly available sources or World Bank country teams; and • Publicly available datasets covering relevant ecosystem indicators including Global Findex, ID4D Administrative Dataset, World Development Indicators, and the IMF Financial Access Survey, among others, and as specified in the analysis. Throughout the note, number of beneficiaries refers to the estimated number of individuals living in households that received social assistance payments from vertically or horizontally expanded programs during COVID-19. In some cases, assumptions were made where the population covered by different programs is known to overlap or where a single program might make payments to multiple beneficiaries within the same household. For more context on these adjustments, see Johnson and Palacios 2022. In the case of payment methods, programs could use one or more method to deliver the social assistance benefits. This could be due to different methods used across regions, across existing versus new beneficiaries, or different payment service providers hired or leveraged. For countries with several programs, different payment methods might also be used across programs. In these cases, countries were catalogued as using digital payments if at least one of the payment methods used was digital. The different payment method categories used are defined as follows: f Digital payments include account-based payments and electronic non-account-based payments. y Account-based payments include fully functional accounts and limited purpose accounts. Either of these can be a mobile money account or traditional account with a financial service provider including banks, SACCOs, cooperatives, and postal banks. • Fully functional account: Benefit is transferred to a recipient’s account, which can be used to withdraw money, make payments, or deposit money. These accounts  can be savings, checking, or basic accounts at a commercial bank, as well as mobile wallets.   29 | The Role of Digital in the COVID-19 Social Assistance Response • Limited purpose account: Benefit is transferred to a recipient’s account that can only be used for a specific purpose. Access to the account or the use of its funds is restricted, with limitations  on how long funds can be stored in the account. The beneficiary cannot deposit money into the  account and  may not be able to make purchases using an associated payment instrument.  These can include pre-paid cards that are topped up by the program, or accounts with limited-purpose debit cards (these do not bear the branding of a major card scheme and can only be used for limited-purpose payments)  y Electronic non-account-based payment: Benefit is transferred to beneficiary via a non-account electronic payment method. The beneficiary accesses their payment using a unique, code- based payment,  one-time password,  single use  prepaid cards, or biometric authentication. Unique-code-based payments are those where the user receives or generates a code (i.e., One-Time-Password) that can be keyed into access points (i.e. ATMs, Points-Of-Sale) to access the funds. No transfer is made into a beneficiary’s personal account.  f Manual payments include those that are made using paper-based methods such as cheques, physical vouchers or physical cash. y Cheques or vouchers: Benefit is distributed to recipients using paper-based payment methods including cheques or vouchers. No transfer is made into a beneficiary’s personal account.  y Cash: Benefit is distributed to recipients in cash, paid out by the line ministry or a third party. No transfer is made into a beneficiary’s personal account.  The Role of Digital in the COVID-19 Social Assistance Response | 30 Endnotes 1  stimate based on payments made as of May 2021. Recipients are measured as estimated individuals living in households that E received a cash transfer, regardless of whether the program targeted individuals or households. 2  igital public platforms can play a critical role for effective public administration, service delivery, and innovation across multiple D sectors. For government-to-person (G2P) programs in particular, systems for digital identity verification or authentication, digital payment systems, as well as structures that facilitate data exchange between these and other sectoral databases and applications, are key elements of a modern G2P architecture. Combined, these platforms can work together as a “stack” to support service delivery and a dynamic digital economy. For more information see Metz et al. 2022. “A Digital Stack for Transforming Service Delivery ID-Payments and Data Sharing” World Bank ID4D. 3  he availability of information across programs and countries varied, resulting in missing values across some indicators. While T all available information was used (i.e. no outliers were removed) some of the descriptive analysis and graphs only cover a subset of countries based on information availability. The number of countries covered is indicated in the respective graphs. 4 Digital payment methods included the use of traditional or mobile money accounts, as well as one-time passwords.  5 J  ohnson and Palacios (2022) provide a number of country cases to illustrate this point. The same paper provides evidence based on more than 80 countries that the use of administrative databases and digital identifiers is correlated with higher coverage of COVID-19 related cash transfers even after taking into account the economic impact of the pandemic. 6 Another example is Haiti although data collection is being contracted out. 7 Cho et al. (2021) COVID-19 G2P Cash Transfer Payments Case Study: Philippines. G2Px Initiative. 8 n at least 20 countries, multiple payment methods were used either within the same program or across different programs in I the country. 9 Countries included in this analysis that had available payment information. Programs that had not yet launched as of May 2021 and programs that only provided advance payment to beneficiaries were excluded. In total, the information from 178 programs across 85 countries is referenced in the text. All of these are developing countries with the exception of seven high-income countries eligible for IBRD lending: Chile, Mauritius, Panama, Romania, Seychelles, Trinidad and Tobago and Uruguay.  ccount-based payments are a starting point for financial inclusion. The pathway between owning an account and meaningful 10 A financial inclusion, through which an individual actively uses that account to meet daily and future financial needs, is not automatic. However, account ownership is widely accepted as a critical first step toward this level of financial inclusion.  his figure overestimates recipients of account-based payments since at least 45 million are recipients of programs that used 11 T accounts in addition to other payment methods. Even among countries that used only accounts, there are usually exceptions made for remote areas or specific populations which were not considered in the categorization as multiple payment method. 12 Among countries analyzed, 41 are using more than one payment method to deliver social assistance payments, either because the approach varies across different regions/populations, or because the payment method varies across programs within the same country. 13 This includes 13 programs that used both limited purpose and fully functional accounts in their delivery.  DD requirements are set by financial institutions to comply with national regulations for identifying and verifying customer 14 C details in advance of open a financial account. National regulations are heavily informed by guidelines provided by the Financial Action Task Force (FATF), the global standard-setting body charged with anti-money-laundering and countering the financing of terrorism (AML/CFT). CDD requirements include, but are not limited to, those details collected at the time of new account onboarding, commonly referred to as know-your-customer (KYC) requirements. 15 Banca de las Oportunidades 2019 16 Maralunda Consultores 2020  n average, countries using digital payments had 37.5 ATMs, 12.3 commercial bank branches, and 64.5 bank agents per 17 O 100,000 adults versus 16.4 ATMs, 9.7 branches, and 58.7 bank agents per 100,000 adults in countries using manual payments, although the difference in averages is only statistically significant in the case of ATMs. On average, countries using mobile money accounts had 620 mobile money agents per 100,000 adults versus 211 in countries using other payment methods. This difference is statistically significant. Source: 2021 G2Px/ASPIRE administrative dataset and IMF’s latest available Financial Access Survey data per country (more than 80% of countries in the sample reported in 2019). 18 Angola used both traditional and mobile money accounts, while Liberia only used mobile money accounts.  orld Bank. 2021. “O sistema de pagamento utilizado pelo Auxílio Emergencial, a introdução da conta social digital e a 19 W bancarização de mais de 100 milhões de pessoas em 9 meses.”  TMs were only used in the cases of Guatemala and Morocco, in addition to agents and branches. In Honduras only branches 20 A and agents, including retail stores where beneficiaries could use their funds to make purchases, were used. 31 | The Role of Digital in the COVID-19 Social Assistance Response  ina Alsali et al. “Lockdown but not Shutdown.” JoPACC. October 2020. https://www.jopacc.com/ebv4.0/root_storage/en/ 21 D eb_list_page/covid-19_report_english.pdf 22 https://www.bcp.gov.py/indicadores-y-datos-de-bancarizacion-i947 23 CAIXA. Relatório da Administração 4T20. https://www.caixa.gov.br/Downloads/caixa-governanca/Relatorio_da_Administracao_4T20. pdf Maralunda Consultores 2021 24   aralelo32.”El Coronavirus adelantó la meta de inclusión financiera en Argentina.” September 3 2020. https://paralelo32.com. 25 P ar/el-coronavirus-adelanto-la-meta-de-inclusion-financiera-en-argentina/ 26 Glynn-Broderick et al. 2022. “Monitoring Digital Financial Payments of Cash Transfers in the Philippines.” G2Px Initiative. 27 Maralunda Consultores 2021  orld Bank. 2021. “O sistema de pagamento utilizado pelo Auxílio Emergencial, a introdução da conta social digital e a 28 W bancarização de mais de 100 milhões de pessoas em 9 meses.”  he cash-out freeze period varied by payment round, with each round being longer. For more information see World Bank 29 T (2021).  mma Wadie Hobson and Silvia Baur-Yazbeck. 2021. “The Benefits of G2P Choice: Zambia Leads the Way.” CGAP. April 1 2021. 30 E https://www.cgap.org/blog/benefits-g2p-choice-zambia-leads-way  oonyoung Cho et al. 2021. “COVID-19 G2P Cash-Transfer Payments Case Study: Philippines.” World Bank. May 2021. https:// 31 Y documents1.worldbank.org/curated/en/368351625163110870/pdf/COVID-19-G2P-Cash-Transfer-Payments-Case-Study- Philippines.pdf Glynn-Broderick et al. 2022. “Monitoring Digital Financial Payments of Cash Transfers in the Philippines.” G2Px Initiative. 32   he clawback clause was initially 90 days, then further expanded to 120. Exceptions were made especially for Bolsa Família 33 T beneficiaries, indigenous registered on Cadastro Único and residents of Acre and Amazonas for whom the period before clawback was 270 days. More information in World Bank (2021).  reta Bull et al. 2021. “Empowering Women: Send Social Assistance Payments Directly to Her.” World Bank Blogs. March 5 2021. 34 G https://blogs.worldbank.org/voices/empowering-women-send-social-assistance-payments-directly-her  he Pradhan Mantri Jan Dhan Yojana (PMJDY) program is a national Financial Inclusion program that aimed for every unbanked 35 T adult to have an account.  or more information on the PMJDY program see https://www.cgdev.org/sites/default/files/social-assistance-and-information- 36 F initial-phase-covid-19-crisis-lessons-household-survey.pdf and https://egc.yale.edu/reaching-indias-poorest-women-covid-19-relief. n Tunisia, payments were directed to men since they were the default head of household. However, a new government decree 37 I allows women to apply for the cash transfer program.  country’s G2P architecture encompasses all systems, infrastructure, regulations, policies, and design choices that enable and 38 A characterize the end-to-end delivery of G2P payments. A modern G2P architectures is one that leverages shared infrastructure and systems, and delivers payments that provide a choice for recipients, prioritizing their convenience and designing with their needs and protection in mind to achieve the long-term development outcomes of financial inclusion and women’s economic empowerment, as well as government-wide fiscal savings. A modern G2P architectures is one that leverages shared infrastructure and systems, and delivers payments that provide 39  a choice for recipients, prioritizing their convenience and designing with their needs and protection in mind to achieve the long-term development outcomes of financial inclusion and women’s economic empowerment, as well as government- wide fiscal savings. The Role of Digital in the COVID-19 Social Assistance Response | 32 worldbank.org/g2px 33 | The Role of Digital in the COVID-19 Social Assistance Response